Very low birth weight (VLBW) infants can be subjected to oxidative stress in the course of intensive care. We measured 8-hydroxydeoxyguanosine (8-OHdG), a biomarker of oxidative stress, and estimated the degree of oxidative stress in such infants. We also examined if the administered oxygen was related to oxidative stress. Urine samples of 50 Japanese VLBW infants [birth weights: 956.3+/-277.6g, and gestational ages: 28.0+/-2.6 weeks (mean +/- SD)] were collected on various postnatal days and 8-OHdG levels were determined using an ELISA kit. Sixteen term infants served as normal controls. As body weights at sampling increased, the average levels of urinary 8-OHdG decreased. 8-Hydroxydeoxyguanosine levels were: infants under 1000g, 29.5+/-16.4 micromol/mol creatinine (n = 24); 1000-1500g, 23.8+/-14.9 (n = 12); over 1500g, 16.1+/-8.5 (n = 14); and control, 10.9+/-7.2 (n = 16). Significant differences were found between <1000g group and > or = 1500g group (p = 0.0030), <1000g group and control (p < 0.0001), and 1000-1500g group and control (p = 0.0108). Also as postconceptional age at sampling increased, the average levels of 8-OHdG decreased. 8-Hydroxydeoxyguanosine levels were: infants before 252 days (36 weeks) of postconception: 27.4+/-15.5 micromol/mol creatinine (n = 34); after 252 days, 18.2+/-12.5 (n = 16). Differences between <252 days group and control (p < 0.0001), and <252 days group and > or = 252 days groups (p = 0.0253) were statistically significant. Among the three groups based on ambient oxygen concentration (21%, 22-29%, and > or = 30%) there was no significant difference (p = 0.417). The more premature the infants were, the more intense was the oxidative stress, hence, it is the prematurity rather than the administered oxygen which causes oxidative stress in VLBW infants. Drury et al. ["Urinary 8-hydroxydeoxyguanosine in infants and children" Free Radic. Res. 28 (1998) 423-4281 measured urinary 8-OHdG of 28 infants (24-40 weeks gestation) and found no gestation or birthweight related differences. This discrepancy seemed to be because of difference in birth weights and sampling period of the subjects.